Tyler Finds Overcoming Bias Is important, For Others
On Aug. 15, Tyler Cowen didn’t think much of overcoming bias, at least in himself:
How important is overcoming bias? …. Biased estimators are one problem but not the only problem. There is also insufficient data, lazy researchers, inefficient estimators, and so on. Then I don’t see why we should be justified in holding a strong preference for overcoming bias, relative to other ends.
On Sept. 9 Tyler, thinks others should try hard to overcome bias in policy stories:
Wherever there are problems, people look for villains. The subprime mortgage crisis is a case in point. … And since every villain must be punished, the Federal Reserve is being attacked as "bailing out the speculators." … But financial markets rarely fit into simple moral narratives, and much as these stories may comfort many of us, they are not a good guide to understanding financial policy. …
Fed watchers should resist the tendency to put all events into a simple or a morally plausible narrative. Monetary policy is a largely technical subject, and its ups and downs don’t usually fit into the kinds of emotion-laden stories that human beings apply to daily life. The "us versus them" tag registers in human memory, but monetary policy is not always or even usually about moral issues. As Freud famously noted, sometimes a cigar is just a cigar.
Yup, other people sure need to overcome their biases Seems Tyler did fall for the confident proposer bias, by assuming we place a huge priority on overcoming bias. But he (and Arnold) never responded to my request for clarification, so I can’t really tell.
There are two reasons why you might think you have to keep moving towards the other side: Aumann-like theorems and the heuristic that says "we're all horribly biased, and some of those biases are reduced by averaging".
If those on the other side aren't adjusting their opinions at all, then either they aren't ideal Bayesians (shock!!) or they aren't convinced that you're an ideal Bayesian (shock!!!), and in either case the theorems that say you should both keep moving towards one another don't apply.
As for the heuristic, I think it embodies a defeatist attitude. Moving towards the average is doubtless usually a good start, especially if you simply don't know what biases might be motivating you, but surely one can do better. But there's been much discussion of this here already...
I think your focus is a little confusing, because sometimes it appears you are talking about our beliefs, such as whether I should amend my pro-choice stance based on the pro-life side. Other times you seem to be looking at it the way a judge or referee looks at two advocates, each with biases. I think these two cases involve very different issues, though in your statements you seem to float from judging 'others' beliefs, and how they should amend them, to judging our own beliefs.
But fundamentally, the practical implications you seem to be implying (and perhaps my perception is incorrect) simply don't sit well. How do you avoid the infinite regress such as in Stokey-Milgrom's no trade theorem, where you start with your beliefs, and after considering "common Knowledge", my priors, the priors of others, there is no trade. Similarly, considering the biases of others and myself, why should I ever deviate from the norm belief without arrogantly or ignorantly not taking into account others? If an issue can be ranked ordinally, if I believe it should be 1.3, others believe it should be a 10, I adjust to say 1.4 tomorrow. Now assume they haven't changed their beliefs? Do I adjust again towards their belief, to 1.5, until I'm finally in agreement? At what point can I just stop and say, "I believe 2, you believe 10, I understand and disagree"? You seem to be implying that I am irrational if I ever stop moving towards the other side of the debate as long as there are people and arguments on the other side. Its a little like relativism, where you meet the paradox of toleration because you tolerate the intolerant, who then create the intolerant society you hate, but who are you to judge? Beliefs fundamentally come down to a paradox, like wondering about the logic of induction, or the consistency and completeness of arithmetic. Curious, but mainly for college bong sessions.